Fe and Mn

In situ and laboratory treatment tests for
lowering of excess manganese and iron
in drinking water sourced from river–
groundwater interaction
Galip Yuce & Candan Alptekin
Environmental Earth Sciences
ISSN 1866-6280
Environ Earth Sci
DOI 10.1007/s12665-013-2343-x
1 23
Your article is protected by copyright and
all rights are held exclusively by SpringerVerlag Berlin Heidelberg. This e-offprint is
for personal use only and shall not be selfarchived in electronic repositories. If you
wish to self-archive your work, please use the
accepted author’s version for posting to your
own website or your institution’s repository.
You may further deposit the accepted author’s
version on a funder’s repository at a funder’s
request, provided it is not made publicly
available until 12 months after publication.
1 23
Author's personal copy
Environ Earth Sci
DOI 10.1007/s12665-013-2343-x
ORIGINAL ARTICLE
In situ and laboratory treatment tests for lowering
of excess manganese and iron in drinking water
sourced from river–groundwater interaction
Galip Yuce • Candan Alptekin
Received: 13 November 2012 / Accepted: 21 February 2013
Ó Springer-Verlag Berlin Heidelberg 2013
Abstract Manganese and iron are essential nutrients at low
doses. However, long-term exposure in high doses may be
harmful. Human activities and natural sources are responsible for manganese and iron contamination in water. This
study aimed to investigate the source of high manganese and
iron contents in shallow groundwater, where the groundwater is used for drinking purpose, and to decrease the excess
manganese and iron from shallow groundwater. Based on the
on-site analytical results taken from the wells in the study
area, iron contents in water samples varied between 30 and
200 lg/L, which were under the allowable limits of Turkish
Drinking Water Standards (TDWS) (Water intended for
human consumption, http://rega.basbakanlik.gov.tr/eskiler/
2005/02/20050217-3.htm, 2005). However, manganese
levels varied from 30 to 248 lg/L, which some of them are
higher than the allowable limits of TDWS (Water intended
for human consumption, http://rega.basbakanlik.gov.tr/eskiler/
2005/02/20050217-3.htm, 2005) (50 lg/L), and EPA (http://
water.epa.gov/drink/contaminants/index.cfm, 2006) (50 lg/L).
The source of excess manganese is originated mainly from
geogenic source besides Porsuk River interaction in shallow
aquifer in the specific section of the study area. To decrease
high manganese content from the well water and reservoir
water, laboratory and in situ treatment tests were applied.
Among these tests, chlorination, associated with filtration
(by fine sand, active carbon and zeolite) and the use of
G. Yuce (&)
Department of Geological Engineering, Eskisehir Osmangazi
University, Meselik, 26040 Eskisehir, Turkey
e-mail: [email protected]
C. Alptekin
Department of Geological Engineering, Bulent Ecevit
University, Zonguldak, Turkey
different filtration procedures by cation exchange resin were
determined as the most effective methods, which was not
previously applied on-site as a combinative approaches to
reduce the excess manganese in water.
Keywords Drinking water Reduce excess manganese
and iron Filtration Groundwater
Introduction
In our century, about one-third of the world’s population
depends on groundwater to meet their drinking and irrigation water demands (Alley et al. 2002). Alluvial aquifer
is the main source of drinking water for many European
countries (Kedziorek and Bourg 2009). Many shallow
wells have been drilled near river bank due to having high
infiltration rate into alluvial aquifer. However, as reported
in the literature, groundwater becomes undrinkable due to
many different reasons (e.g., land degradation, deforestation, poor agricultural activities, overpopulation, inaccurate
exploitation, and management of the water resources).
Environmental pollution due to natural (including geogenic) or anthropogenic reasons occurred in surface water
resources is relatively more common in proportion to
groundwater (May et al. 2011; Petitta et al. 2011; Rao et al.
2012; Reddy 2013; Choi et al. 2013), this makes groundwater an attractive source for drinking water. However,
water usage and management in upstream area impact
water quality and quantity in downstream area. Hence, the
protection of the quality and quantity of freshwater
resources gains a great importance to sustain its utilization
(Kumar et al. 2010). In this respect, river–groundwater
interaction can alter water quality and quantity by reciprocally (Mohammadi 2009; Gao et al. 2010; Baalousha
123
Author's personal copy
Environ Earth Sci
Fig. 1 Location map of the study area
2012; Edet et al. 2012; Delfs et al. 2012; Borrelli et al.
2012; Garcia et al. 2012).
The shallow groundwater in the study area is exploited
to meet domestic water for the Meselik Campus of Eskisehir Osmangazi University. The campus is located in
close proximity to the course of the Porsuk River (Eskisehir, Turkey). The most common pollutants in the Porsuk
River, which also cause deterioration of groundwater
quality, are mainly manganese and iron as much as nitrogen, phenol, and trace elements (Yuce et al. 2006). Previous research indicates that the Porsuk River interacts with
the shallow groundwater system (Kacaroglu 1991; Kacaroglu and Gunay 1997; Yuce et al. 2006). As a result of this
interaction, shallow groundwater is widely contaminated
by the Porsuk River (Yuce 2007). The reductive dissolution
of Mn and Fe oxides has caused increase in the Fe and Mn
concentrations in groundwater of the shallow aquifer.
Consequently, high dissolved manganese and iron in
drinking water provided from the shallow wells may cause
health problem. For this reason, reducing the manganese
and iron concentrations in the shallow groundwater is the
essential.
123
Manganese and iron are the most common trace elements
in the aquatic environment. But their concentrations in
drinking water are varied in many Guidelines. In this respect,
the allowable upper limit of iron content in drinking water is
200 lg/L according to Turkish Drinking Water Standards
(TDWS 2005) and Guidelines for European Community
Drinking Water Standards (GECDWS 1998) while it goes up
to 300 lg/L based on Guidelines for Canadian Drinking
Water Quality Standards (GCDWS 1996), and Environmental Protection Agency (EPA 2006). For manganese, this
limit is 50 lg/L according to TDWS (2005), GECDWS
(1998), EPA (2006), and GCDWS (1996).
Manganese has usually a divalent ion (Mn2?) in igneous
rocks (Bell 1998). Biotite and hornblende are common
ferromagnesian minerals in igneous rocks besides pyroxenes, amphiboles; magnetite and olivine are other sources
of iron (Appelo and Postma 1992) in rock-forming minerals. Iron and manganese oxyhydroxides may dissolve in
surface waters as a result of photo-reduction (Drever 1997).
Chemical and bacterial reduction of Mn(IV) to Mn(II) has
been seen in groundwater with low oxygen content (Jaudon
et al. 1989). The occurrence of the manganese in suspended
particles, sediments, and water resources (lakes, rivers, and
groundwater) was reported widely (Drever 1997; Berbenni
et al. 2000; Barber et al. 1995).
Oxygenated water has lower iron and manganese contents because they form insoluble compounds in water.
Generally, uncontaminated surface and groundwater have
higher dissolved oxygen content, so iron and manganese
are found insoluble forms in water. In surface water, iron
and manganese are most likely trapped within suspended
organic matter particles. Under anoxic conditions, iron and
manganese remain in their soluble forms of Fe2? and Mn2?
(Domenico and Schwartz 1990; Appelo and Postma 1992;
Fetter 1999). Redox condition in water is mainly characterized by two parameters, redox potential (Eh) and dissolved oxygen (DO). Thus, the oxidizing and reducing
conditions are measured as a function of Eh and DO.
Higher Eh values correspond to more oxidizing conditions
while lower values correspond to more reducing conditions. DO concentration generally is expressed in ppm or %
saturation in related with Eh and pH, by the following
reaction and equation:
H2 O ¼ 2Hþ þ 1=2O2 ðaq Þ þ 2e
ð1Þ
Eh ¼ E0 0:059 pH þ 0:01475 log ½O2 ð2Þ
On the other hand, a recent study (Kedziorek and Bourg
2009) suggests that the electron trapping capacity of
dissolved oxygen and nitrate is a better indicator for
manganese and iron dissolution in alluvial aquifers near
river.
Author's personal copy
Environ Earth Sci
Fig. 2 Sampling location and
geological map of the study area
Many conventional and improved treatment techniques
were searched and applied to decrease or remove iron
and manganese in water are based on physico-chemical
reactions such as usage of manganese greensands or
intense aeration and chemical oxidation or utilization of
polyphosphate, ion exchanger (resin or zeolite), microbiological removal, and chlorination (oxidation) plus filtration (Shavandi et al. 2012; Tarasevich et al. 2012; Dubinina
et al. 2012; Van Halem et al. 2012; Farnsworth et al. 2012;
Tang et al. 2012; Fu et al. 2011; Carriere et al. 2011). All of
these treatment techniques are useful, but most of them
certainly reproduce some other quality problems. For
instance, the application of manganese greensand method
(MGM) (http://www.birmfilter.com/index.asp?p=s10)
requires regeneration process. Potassium permanganate is
used in the regeneration process. Since the potassium permanganate is in pink color, it may cause pink stain in water
unless the water is cleaned by the filtration membrane.
Therefore, MGM is not applicable without an extra process
for removing the excess manganese from water. A regular
backwashing is also essential for effective filter performance
in MGM. Another common approach to reduce excess
manganese in water is aeration of water. In order to obtain a
full achievement in the aeration technique, after abstracting
of groundwater from the well, water must gush onto the
surface at the stair-shaped plates as a first step pooling to
provide full saturation by air. Following this, the aerated
water must be exposed to a second pooling by over-flowing
to collect the water before introducing it to a water reservoir,
which mostly consists of a closed system.
In this study, the most efficient methods to reduce
excess manganese and iron in water without undesirable
remnants were searched and tested as in situ and laboratory
experiments by the innovative combinations of different
individual treatment techniques were performed for the
improving of water quality in the study area.
Materials and methods
Location and geological setting
The study area (6 km2) is located in the northwest of the
Central Anatolian Region of Turkey, in the Eskisehir
Province (Fig. 1).
The basement in the study area is in Triassic age consisting of schist, limestone, and ophiolitic mélange that
overlain by sedimentary (conglomerate, marl, and claystone) and volcano-sedimentary units (agglomerate and
basalt) of Eocene, Miocene, and Pliocene (Gozler et al.
1996). Pleistocene and Holocene units are composed of old
and young alluvium, respectively. The thickness of young
alluvial deposits varies from 5 to 15 m in the wells opened
in the study area with 14 and 70 m depths. Old and young
alluvial deposits form the main aquifer units. Young alluvium outcropped around Porsuk River comprises loose
123
Author's personal copy
Environ Earth Sci
Table 1 Water analyses results (September and October 2009) (samples were taken after long period of pumping)
Temp
(oC)
TDS
(mg/L)
DO
(mg/L)
Eh
(mV)
pH
Sample No
MCL for Turkish Drinking Water Standards
(TDWS), 2005
MCL for European Community Drinking Water
Standards (GECDWS), 1998
MCL for Canadian Drinking Water Quality
Standards (GCDWS), 1996
MCL according to the EPA, 2006
Sampling
Sampling Locations
from
specific
depth (m)
1 The active well
Filtered by
5
2 The recently drilled
well (spare
(0.45 µm)
well) (inactive)
10
SC
(µS/cm)
In-situ
colorimetric
analyses
results (µg/L)
200
200
50
50
300
50
Analyses results in
EMMLs
(mg/L)
50
300
50
Fe
Mn Fe/Mn
0.5
10
NH4
NO3
NO2
13.7
14.4
767
670
0.50
0.44
4.6
7.3
375
365
7.5
8.0
40
120
248
97
0.15 <0.05
1.23
nd
7.06
12.48
0.143
0.878
nd
nd
12.0
nd
nd
682
nd
nd
0.44
nd
nd
2.0
nd
nd
175
nd
7.4
30
30
160
108
78
164
0.27
1.15
0.98
nd
nd
nd
15.70
17.65
1.86
1.504
1.452
0.003
4 Well
5 Well
6 Spare well-27279 (inactive)
nd
nd
13.4
15.2
15.3
nd
nd
694
1441
919
nd
nd
0.45
0.94
0.60
2.2
nd
3.8
5.8
6.3
181
nd
243
368
366
nd
nd
8.1
7.7
7.9
100
180
100
200
90
160
152
30
32
52
0.63
nd
1.18
nd
nd
3.33
6.25 <0.05
1.73 <0.05
2.92
1.60
3.54
36.50
25.35
0.007
0.025
0.002
0.010
0.011
7 Well-27278
8 Porsuk River
9 Reservoir water
Tap water
15.5
15.5
nd
nd
924
618
nd
nd
0.62
0.37
nd
nd
4.3
7.8
nd
nd
294
nd
nd
nd
8.1
7.5
nd
nd
nd
50
30
110
67
212
148
104
nd
0.24
0.20
1.05
22.75
5.66
5.38
7.38
0.011
0.132
0.127
0.016
3 Spare well
(inactive)
Filtered by
(0.45 µm)
17
14
20
31
<0.05
<0.05
<0.05
<0.05
nd not detected, nf not filtered, EMMLs Eskisehir Metropolitan Municipal Laboratories
gravel, sand, silt, and clay materials and their mixture.
Alluvial deposits are underlain by ophiolitic mélange. The
alluvial aquifer is unconfined character in the study area.
The yields of the wells range from 18 to 90 m3/h. The
study area contained five actively operated wells, which
provide drinking water to the University, and two nonoperative spare-wells. The active well, No. 1 and the sparewells (inactive), No. 2 and No. 3 are very close (within
50 m distance) to the Porsuk River. The thickness of recent
alluvium in the well No. 2 is 8 m. The depth of this well is
20 m drilled nearby the Porsuk River. The recent alluvium
in this well is underlain by ophiolitic mélange in 12 m
thick at the bottom.
Selection of sampling locations
In the dry period of 2009, water samples were taken from
the water-wells and water reservoir (Fig. 2). The dry period
was selected for sampling period because worst-case-based
water quality mostly occurs in this period due to the lack of
dilution by precipitation (rainfall, etc.). In total, 14 water
samples were collected from 7 water-wells in the various
depths, Porsuk River, reservoir, and tap water. All sampled-wells drilled in the shallow phreatic aquifer are along
the Porsuk River. The locations of the sample wells were
specified according to Global Position System (GPS).
123
Water levels in wells were measured by a water level
measurement unit (hydrometer) and hydraulic heads
changed from 801 to 796 m above sea level (a.s.l.) in
direction from well No. 1 to well No. 4 while water level in
the Porsuk River varied from 803 to 802 m a.s.l. in the
sampling section of the study area (i.e., hydraulic head
measurements were only available for the wells No. 1
through No. 4 since the rests were equipped with submersible pumps).
Sampling and analytical procedure
Water samples were collected from operating and nonoperating well and analyzed in situ for manganese and iron
contents. Water samples from the spare (non-operating)
wells were collected under stagnant condition by the deep
sampling device (hydro-bios) and a portable-pump (Power
Pack PP1 with Waterra inertial pump) under the operating
(active) condition in varied depths to identify any difference between stagnant and active conditions regarding
manganese and iron content. In situ iron and manganese
analyses were carried out based on the FerroVer (AccuVac
Ampules) and PAN methods, respectively. The detection
limits of these methods are 30 and 20 lg/L for dissolved
iron and manganese, respectively. An ascorbic acid reagent
was used initially to reduce all oxidized forms of
Author's personal copy
Environ Earth Sci
Table 2 Major and minor ion analysis of the active university’s well
(No. 1)
Ion
Concentrations (meq/L)
Na
0.91
K
0.02
Ca
3.11
Mg
4.44
Total cation
8.48
HCO3
6.56
SO4
Cl
0.92
0.70
NO3
0.02
Total anion
8.20
Ion balance error %
1.67
Trace elements
(lg/L)
Ni
1.48
Al
9.26
Fe
40
Mn
248
Cr
2.08
As
7.04
Fig. 5 Dissolved Mn versus pH
Fig. 6 Dissolved Fe versus dissolved oxygen (DO%)
Fig. 3 Eh versus dissolved oxygen (DO%)
manganese to Mn2?. An alkaline-cyanide reagent was
added to mask any potential interference. PAN Indicator
was then added to combine with the Mn2? to form an
orange-colored complex. Since unfiltered water samples
caused a change in the manganese and iron contents by
artificial dissolution in suspended sediments in water, the
collected water samples were filtered by 0.45 lm filter
before analyzing. Detailed procedure for in situ colorimetric analysis can be obtained from Hach-DR/890, 2007.
The NH3, NO2, NO3, and NH4 analyses were performed in
Eskisehir Metropolitan Municipal Laboratories (EMMLs).
Major ions and trace elements were analyzed by Ion
Chromatography in the EMMLs. The specific conductance
(SC), water temperature (°C), total dissolved solid (TDS),
pH, and Eh values were measured in situ by YSI-556
Handheld Multi-parameter Instrument. Error limits (accuracy) for water temperature, Eh, SC, and pH were
±0.15 °C, ±20 mV, ±0.5 and ±0.2 %, respectively. Trace
element analyses were performed by ICP-OEM.
Experimental studies
Fig. 4 Dissolved Mn versus Eh
The iron contents of the water samples were lower than the
maximum allowable limit. Thus, much effort has been
spent on decreasing the manganese concentration in the
water to a safe drinking water quality. To reduce the excess
123
Author's personal copy
Environ Earth Sci
Fig. 7 Dissolved Mn versus dissolved oxygen (DO%)
Fig. 9 Dissolved Mn versus NO3
Fig. 8 Dissolved Fe versus NO3
manganese in water, laboratory, and in situ experimental
methods were applied. For the laboratory tests, the sample
taken from operating well No. 1 was subjected to four
different experiments: (1) one part of the samples was
rinsed under aeration (circulation of raw water through
cascades to air) and the other part was only aerated without
rinsing, the manganese concentrations of both parts in the
sample were measured after 18 h of aeration, (2) another
part of the sample from the same well was filtered by a
0.45 lm and the other part was passed through a smaller
sized filter by 0.2 lm. Both filtered water samples were
aerated separately, (3) after application of above-mentioned filtration procedures, the ozonation was performed
to the both filtered water samples instead of aeration, (4) in
the last experimental method, two filtered water samples
were exposed to chlorination. At the end of each method,
the manganese contents were measured.
In addition to the laboratory applications, on-site manganese lowering methods were also tested for the same
well water sample (i.e., well No. 1). For this purpose, four
methods were applied using on-site tests: (1) the combination of chlorination and filtration (sand and active charcoal were tested), (2) the combination of chlorination and
filtration (sand, active charcoal, and 0.2 lm extra filter
were tested), (3) the combination of chlorination and
123
Fig. 10 Variations of manganese and iron concentrations depending
on the depth in well No. 3 a before and b after pumping
filtration (sand, active charcoal, and zeolite were tested), and
(4) the combination of sand and filtration (only cation
exchange resin was used). Water from the well No. 1 was
retained for 45 min in aeration following chlorination and
then it was filtered. Same procedure was also repeated for
water samples collected from the spare wells (No. 2 and No.
3). The water samples from these wells taken at three different
depths were analyzed in the laboratory to evaluate the
extension of oxidative/reductive environment in the shallow
aquifer condition. At the end of each test, the manganese
content was measured by colorimetric technique.
Results and discussion
The analytical results of the water samples are given in
Table 1. The minimum and maximum concentrations of
Author's personal copy
Environ Earth Sci
Fig. 11 Manganese coating on the pipes and pump in the active well No. 1
iron and manganese vary from 30 to 200 lg/L and from 30
to 248 lg/L, respectively. The results show that the
manganese content is higher than the allowable limits for
drinking standards. Iron contents of the groundwater samples are generally lower than the maximum allowable value
recommended by TDWS (2005), GCDWS (1996), EPA
(2006), and GECDWS (1998). However, the manganese
content in the water samples is quite higher than the
allowable limit (50 lg/L) for 12 water samples (Table 1)
out of 14. Particularly, the water sample taken from the
active well (No. 1) nearby the Porsuk River has the highest
manganese content (Table 2).
The manganese and iron contents in the samples taken
from various depths of the spare well No. 3 show no
remarkable change in the depth. Thus, this indicates that a
reductive environment prevails in the environment of well
No. 3 and relative decrease occurs in the Fe/Mn ratios
(Barber et al. 1995) and nitrate values in comparison with
the environment of wells No. 4 and No. 5 (Table 1). This is
likely to be the result of nitrate serves as electron acceptor
for organic matter in anoxic condition.
The in situ measurements of Eh and DO% show a linear
correlation (Fig. 3). By reviewing the analytical results, an
inverse relationship between Eh and dissolved Mn is
observed (Fig. 4), except for sample 1. In other words,
dissolved Mn decreases with oxidation due to conversion
of Mn to the insoluble forms. Thus, there is an inverse
correlation between higher oxide forms of manganese with
an increase in pH (Fig. 5), which is consistent with the
theoretical expectations. Samples from well No. 2 in 5 m
depth, well No. 3 in 14 and 20 m depths were used in
Figs 3, 4, 5, 6, and 7. Since there is no simple trend
between Fe and Mn against DO% (Figs. 6, 7), DO may not
be a significant electron acceptor for them. However, NO3
seems to overtake this role by nitrification process instead
of DO (Figs. 8, 9). The occurrence of nitrification process
(Fig. 9) in shallow aquifer system is related with the
increasing thickness of alluvium. Eh, DO, and NO3 values
in water samples from the wells No. 3 and No. 4, relatively
low. A corresponding increase is observed in dissolved iron
and especially manganese. According to the obtained data,
sample 1 (from well No. 1) showed a relative decrease in
DO and NO3 concentrations, but Eh and manganese concentration increase, which may be resulting in a reductor
role of iron for manganese (Appelo and Postma 1992). In
contrast to the sample 1, sample 5 (from another well)
showed an increase in iron concentration and decrease in
manganese concentration which may indicate an oxidative
role of manganese for iron (Kedziorek and Bourg 2009).
The similar increase in manganese concentration and
decrease in NO3 concentration in well No. 1 and the Porsuk
River may result in the bank infiltration from the Porsuk
River into shallow aquifer. Thus, recharge/discharge condition between the Porsuk River and shallow aquifer can
create the oxidation/reduction process in the aquifer system. Porsuk River water is originated from Porsuk dam’s
lake located in the upstream area with ca. 20 km distance
to the sampling site. Mn and Fe contents in water from
Porsuk dam’s lake are lower (average 70 lg/L–RDMFE
2005) up to the sampling site. After the outlet of Porsuk
123
Author's personal copy
Environ Earth Sci
Table 3 Laboratory experiments for the removal of manganese and iron from water
Element
Unit
Untreated water sample
from well No. 1
Aerated 18 h
Aerated
Rinsed
Stagnant
0.45 lm
filter
0.2 lm
filter
Ozonated by 0.7 ppm
Chlorinated by 1.8 ppm
0.45 lm
filter
0.45 lm
filter
0.2 lm
filter
0.2 lm
filter
Manganese
lg/L
248
51
248
38
28
51
33
71
27
Iron
lg/L
40
10
20
20
10
20
10
20
10
dam’s lake, Fe and Mn levels increase in the river water
most likely to be the result of groundwater recharge having
high Fe and Mn from Eocene basalts and ophiolite caused a
reductive environment in shallow groundwater at the
upstream of sampling site. Hence, DO and NO3 concentrations decreased in the well samples No. 1, No. 3, and No.
4. Indeed, since the river is in contact with atmosphere, Mn
and Fe are expected to be easily oxidized and converted to
oxide and oxy-hydroxide precipitates. However, presence
of Mn and Fe in Porsuk River water (sample 8) is likely
that the river water has not had enough time to fully
oxidize the all Mn and Fe influx for this sampling section.
Fe/Mn ratio was low in the samples 1, 8, and 9. In the
transition zone (anoxic–oxic), Fe/Mn ratio especially
decreases with the precipitation of Fe as it was identified
for these samples.
Fig. 12 Manganese and iron
concentrations in water (from
well No. 1) following laboratory
treatment methods
123
When Fe2? is oxidized to Fe3? and combined with an
appropriate anion, it precipitates as iron oxyhydroxide
(FeOOH) in water and if manganese co-exists with iron in
water, it precipitates as the hydroxyl of Mn3? after the
precipitation of Fe3?. This is because oxidation of iron
faster than manganese providing that existing of iron bacteria, which catalyze the oxidation reactions under certain
conditions of pH and Eh (WHO 2004). Consistent with this,
we observed slight lesser amount of iron in water of the
well No. 3 (non-operated for a long time) due to exposure
in the oxidative condition by pumping and the variations of
manganese and iron contents in different depths before and
right after pumping were given in Fig. 10 (concentrations
shown in Fig. 10 were measured before and right after
pumping, therefore they are different than that of given in
Table 1). Interestingly, manganese concentration increased
Author's personal copy
Environ Earth Sci
Table 4 In situ manganese removal experiments according to the analyses results
Untreated water
sample from well
No. 1
Manganese
lg/L
248
Sodium
mg/L
20.9
Well water retained for
45 min followed by
chlorination
Sand ? Active Carbon
Well water retained for
45 min followed by
chlorination
Sand ? Active
Carbon ? Zeolite
Well water retained for
45 min followed by
chlorination
Sand ? Active
Carbon ? 0.2l filter
210
25
72
Raw Well
Water
Sand ? Resin
0.00
163
nd not detected, nf not filtered, EMMLs Eskisehir Metropolitan Municipal Laboratories
immediately following pumping and it is most likely as a
result of flushing effect that migrate manganese-rich stagnant water at the deep to the surface by pumping. Furthermore, manganese removal bacteria require completely
oxic environment (Johnson et al. 2009). After oxidation,
water color became brownish due to the precipitation of the
ferric hydroxide and manganese hydroxide in water.
Hence, manganese coatings were observed on the innerwalls of water reservoir as well as on the stainless-steel
pipelines and the submersible pump (Fig. 11).
Since the hydraulic head (801–796 m a.s.l.) in alluvial
aquifer in the south of Porsuk River is lower than that of
Porsuk River (803–802 m a.s.l.), alluvial aquifer is fed by
this particular section. The wells (No. 1, No. 2, and No. 3)
drilled near the Porsuk River have excess manganese
concentration especially due to the river–groundwater
interaction along the section of the study area, except for
the well No. 4. This is because well No. 4 is extracted
groundwater from the deeper aquifer and well equipped
with stainless steel casing, and the annulus between well
wall and casing is sealed up to the aquifer level, to protect
it against surface contamination. Therefore, river water
cannot seep into the well No. 4, but the other wells are
exposed to river influence due to the lack of such a protective construction in addition to lower water level in the
wells compare to the water level in the river (see Sect. 2.2).
Besides river influence, there also exists a lithologic effect
since Fe–Mn-rich Eocene basalt and ophiolitic basement,
outcropped in the northwest of the study area. Thus, the
alluvium contains debris from the ophiolitic basement and
Eocene basalt, which is abundant in manganese and iron
rich minerals.
Results of chemical analyses show that HCO3 is dominant
anion in well water (No. 1) and the Porsuk River (Table 2).
HCO3- and Mn can form MnHCO3 as a result of chemical
interaction. If water is exposed to oxidative condition, then
manganese, sourced from MnHCO3, can transform into
insoluble oxides and this may result in the precipitation of
Mn. When manganese bicarbonate is exposed to oxygen, it
converts into manganese hydroxide which is less soluble
form (Allen and Pelude 2001), and then manganese
hydroxide precipitates as black scale deposits. This is why
the inner-walls of the water reservoir are colored in dark.
To achieve satisfactory result for removing manganese
and iron from water without problematic remnants, as
mentioned in a previous section, four different laboratory
tests were applied: (1) aeration, with and without rinsing,
(2) aeration followed by filtration using two different types
of filters, (3) ozonation, and (4) chlorination following two
different sized filtrations (Table 3). The application of the
second, third, and fourth laboratory tests provided sufficient decrease in the manganese to meet the acceptable
quality for drinking water standards (Fig. 12; Table 3).
On-site experimental methods were the combination of
chlorination and filtration that consisted of the different
materials such as sand, active charcoal, zeolite, and resin.
By application of the on-site tests, manganese concentration of water from the active well No. 1 was decreased
below the allowable limit for drinking water. Particularly,
by using the third and fourth in situ methods, manganese
content was successfully decreased from 248 lg/L to a
lower level than the TDWS (2005) (Table 4; Fig. 13). By
employing the fourth on-site method, the manganese and
iron contents were completely removed from the water.
However, the unique disadvantage of the fourth method
was the increase in the sodium content of water due to the
calcium and magnesium adsorption, besides iron and
manganese. In spite of this, the sodium content was still
under the allowable limit according to TDWS (2005).
In conclusion, our study provided a better understanding
about which methods were the best for reducing of excess
Fig. 13 Manganese concentrations in water (from well No. 1)
following in situ treatment methods
123
Author's personal copy
Environ Earth Sci
manganese in water. As a result, the application of two
methods, namely, the filtration by the combination of sand (4th
in situ method) and resin without chlorination, and the filtration
by the combination of sand, active carbon, and zeolite with
chlorination (3rd in situ method) were the most practical and
effective methods among the four on-site methods.
Osmangazi University and Ahmet Sekerci from Regional Directorate
of State Hydraulic Works (DSI) for their invaluable supports during
field and laboratory studies. The authors would like to extend their
sincere thanks to three anonymous reviewers for their insightful and
constructive comments on this manuscript.
References
Conclusions
We found high manganese content in water samples taken
from the wells opened in shallow groundwater system to
provide drinking water near the Porsuk River. On the basis
of the analytical results of water samples, the manganese
content in eight out of ten were over the allowable limit
according to national and international drinking water
standards. This shallow aquifer system has originally a
reductive character (having low nitrate content), converts
into an oxidative environment by the infiltration from
Porsuk River. This change leads to the alteration of the
dissolved Mn and Fe concentrations in the groundwater.
Thus, the major reason of the excess manganese and iron
contents in shallow groundwater is due to the interaction
between Porsuk River and shallow groundwater system.
Conformingly, groundwater flow patterns, water chemistry
(Yuce et al. 2006; Yuce 2007; Alptekin 2012), temperature,
pH, NO3 data indicate such a mixing from the Porsuk River
into the shallow groundwater system. Moreover, lithology
of alluvial aquifer is another reason for excessive manganese and iron since aquifer composes particularly of mafic
minerals (rich in iron and manganese) associated with
volcanic and ophiolitic rocks outcropped in the study area.
To reduce the excess manganese and iron content in
shallow groundwater, we applied both laboratory and
on-site methods on collected water samples from the wells
in the study area. By the application of the laboratory
methods, such as the coupled processes of filtration and
aeration, filtration and ozonation, and filtration and chlorination, satisfactory results were obtained in reducing
manganese and iron content of the well water even well
below than the allowable limits for drinking purpose.
Filtration by the combination of sand and resin without
chlorination, as well as the filtration with the combination
of sand, active carbon, and zeolite with chlorination were
the most considerable and economic methods among the
four on-site methods to diminish excess manganese and
iron content in the groundwater. Finally, excess manganese
content was decreased to 0 and 25 lg/L by application of
these two methods, respectively.
Acknowledgments The authors are grateful to Hakki Demirtas
from Water and Sewage Authority of Eskisehir Metropolitan
Municipality, Semra Ilhan from Chemistry Department of Eskisehir
123
Allen DM, Pelude K (2001) Dissolved manganese in drinking water
on the Gulf Islands: occurrence and toxicity. Report prepared for
the Islands Trust and the BC Ministry of Environment, Lands
and Parks, p 36
Alley WM, Healy RW, LaBaugh JW, Reilly TE (2002) Flow and
storage in groundwater systems. Science 296:1985–1990. doi:
10.1126/science.1067123
Alptekin C (2012) Examination of surface water and groundwater
interaction by contamination parameters: a case study of Porsuk
River (Eskisehir), Msc. Thesis, Eskisehir Osmangazi University,
Institute of Science, p 110 (in Turkish)
Appelo T, Postma D (1992) Geochemistry, groundwater and pollution, vol 7(1). Balkema, Rotterdam, pp 18–19
Baalousha HM (2012) Modelling surface-groundwater interaction in
the Ruataniwha basin, Hawke’s Bay, New Zealand. Environm
Earth Sci 66(1):285–294
Barber BL, Leenheer AJ, Pereira WE, Noyes TI, Broun KG, Tabor
CF, Writer JH (1995) Organic contamination of the Mississippi
River from Municipal and Industrial Waste Water. US Geological Survey Circular 1133, Reston, Virginia
Bell GF (1998) Environmental geology: principles and practice.
Blackwell Science, ISBN: 0-86542-875-1, 509-511.
Berbenni P, Pollice A, Canziani R, Stabile L, Nobili F (2000)
Removal of iron and manganese from hydrocarbon-contaminated groundwaters. Bioresour Technol 74:109–114
Borrelli N, Osterrieth M, Romanelli A, Alvarez MF, Cionchi JL,
Massone H (2012) Biogenic silica in wetlands and their
relationship with soil and groundwater biogeochemistry in the
Southeastern of Buenos Aires Province. Argentina. Environ
Earth Sci 65(2, Special Issue):469–480
Carriere A, Brouillon M, Sauve S, Bouchard MF, Barbeau B (2011)
Performance of point-of-use devices to remove manganese from
drinking water. J Environ Sci Health Part A Toxic/Hazard Subst
Environ Eng 46(6):601–607
Choi BY, Yun ST, Kim KH, Kim K, Choh SJ (2013) Geologically
controlled agricultural contamination and water-rock interaction
in an alluvial aquifer: results from a hydrochemical study.
Environ Earth Sci 1:203–217
Delfs JO, Blumensaat F, Wang WQ, Krebs P, Kolditz O (2012)
Coupling hydrogeological with surface runoff model in a Poltva
case study in Western Ukraine. Environ Earth Sci 65(5, Special
Issue):1439–1457
Domenico PA, Schwartz FW (1990) Physical and chemical hydrogeology. Wiley, Canada
Drever JI (1997) The geochemistry of natural waters surface and
groundwater environments, 3rd edn. Prentice Hall, Upper Saddle
River
Dubinina GA, Sorokina AYu, Mysyakin AE, Grabovich MY,
Eprintsev AT, Bukreeva VY (2012) Modelling and optimization
of processes for removal of dissolved heavy-metal compounds
from drinking water by microbiological methods. Water Resour
36(4):398–404
Edet AE, Worden RH, Mohammed EA, Preston MR (2012)
Hydrogeochemical processes in a shallow coastal plain sand
aquifer and tidal river systems (Calabar, Southeastern Nigeria):
Author's personal copy
Environ Earth Sci
tracking wastewater and seawater pollution in ground and river
waters. Environ Earth Sci 65(7):1933–1953
EPA (2006). http://water.epa.gov/drink/contaminants/index.cfm
Farnsworth CE, Voegelin A, Hering JG (2012) Manganese oxidation
induced by water table fluctuations in a sand column. Environ
Sci Technol 46(1):277–284
Fetter CW (1999) Contaminant hydrogeology, 2nd edn. Prentice Hall,
Upper Saddle River
Fu J, Shang J, Zhao Y (2011) Simultaneous removal of iron,
manganese and ammonia from groundwater in single biofilter
layer using BAF. In: Shi YG, Zuo JL (eds) Environmental
biotechnology and materials engineering, PTS 1-3, vol 183–185.
Advanced materials research, pp 442–446
Gao XB, Wang YX, Wu PL, Guo QH (2010) Trace elements and
environmental isotopes as tracers of surface water-groundwater
interaction: a case study at Xin’an karst water system, Shanxi
Province, Northern China. Environ Earth Sci 59(6):1223–1234
Garcia MG, Lecomte KL, Stupar Y, Formica SM, Barrionuevo M,
Vesco M, Gallara R, Ponce R (2012) Geochemistry and health
aspects of F-rich mountainous streams and groundwaters from
sierras Pampeanas de Cordoba. Argentina. Environ Earth Sci
65(2, Special Issue):535–545
GCDWS (1996). http://www.hc-sc.gc.ca/ewhsemt/pubs/water-eau/
2010-sum_guide-res_recom/index-eng.php
GECDWS (1998). http://www.lenntech.com/applications/drinking/
standards/eu-s-drinking-water-standards.htm
Gozler Z, Cevher F, Ergul E, Asutay J (1996) Geology of southern
and middle part of Sakarya. MTA Publications no. 9973:97
Ankara (in Turkish)
Hach DR/890 (2007) Colorimeter; for in situ Fe and Mn colorimetric
analysis procedure manual
Jaudon P, Massiani C, Galea J, Rey J (1989) Groundwater pollution
by manganese. Manganese speciation: application to the selection and discussion of an in situ groundwater treatment. Sci Total
Environ 84:169–183
Johnson M, Ratnayaka DD, Brandt JM (2009) Twort’s water supply,
6th edn. ISBN: 978-0-7506-6843-9, p 711
Kacaroglu F (1991) Study of contamination in the Eskisehir Basin,
PhD Thesis, Hacettepe University, Ankara (in Turkish)
Kacaroglu F, Gunay G (1997) Impacts of human activities on
groundwater quality of an alluvial aquifer: a case study of the
Eskisehir Plain (Turkey). Hydrogeol J 5(3):60–70
Kedziorek MAM, Bourg ACM (2009) Electron trapping capacity of
dissolved oxygen and nitrate to evaluate Mn and Fe reductive
dissolution in alluvial aquifers during riverbank infiltration.
J Hydrol 365:74–78
Kumar S, Bharti VK, Singh KB, Singh TN (2010) Quality assessment
of potable water in the town of Kolasib, Mizoram (India).
Environ Earth Sci 61(1):115–121
May R, Jinno K, Tsutsumi A (2011) Influence of flooding on
groundwater flow in central Cambodia. Environ Earth Sci
63(4):151–161
Mohammadi Z (2009) Assessing hydrochemical evolution of groundwater in limestone terrain via principal component analysis.
Environmental Earth Sciences 59(2):429–439
Petitta M, Primavera P, Tuccimei P, Aravena R (2011) Interaction
between deep and shallow groundwater systems in areas affected
by Quaternary tectonics (Central Italy): a geochemical and
isotope approach. Environ Earth Sci 63(1):11–30
Rao NS, Subrahmanyam A, Kumar SR, Srinivasulu N, Rao GB, Rao
PS, Reddy GV (2012) Geochemistry and quality of groundwater
of Gummanampadu sub-basin, Guntur District, Andhra Pradesh,
India. Environ Earth Sci 67(5):1451–1471
Reddy AGS (2013) Evaluation of hydrogeochemical characteristics of
phreatic alluvial aquifers in southeastern coastal belt of Prakasam district, South India. Environ Earth Sci 68(2):471–485
Regional Directorate of Ministry of Forest and Environment
(RDMFE) (2005) The Environment report of Eskisehir Province
for 2004, published by Eskisehir Governorship, p 299 (in
Turkish)
Shavandi MA, Haddadian Z, Ismail MHS, Abdullah N, Abidin ZZ
(2012) Removal of Fe(III), Mn(II) and Zn(II) from palm oil mill
effluent (POME) by natural zeolite. J Taiwan Inst Chem Eng
43(5):750–759
Tang Y, Wu W, He Y, Fu JX, Wang XL (2012) Low-temperature
domestication of an iron and manganese oxidizing bacteria. In:
Li H, Liu YF, Guo M (eds) Sustainable development of urban
environment and building material, PTS 1–4. Advanced materials research, vol 374–377, pp 826–830
Tarasevich YuI, Goncharuk VV, Polyakov VE, Krysenko DA,
Ivanova ZG, Aksenenko EV, Tryfonova MY (2012) Efficient
technology for the removal of iron and manganese ions from
artesian water using clinoptilolite. J Ind Eng Chem
18(4):1438–1440
Turkish Water Standard (TDWS) (2005) Water intended for human
consumption. (http://rega.basbakanlik.gov.tr/eskiler/2005/02/200
50217-3.htm)
Van Halem D, Moed DH, Verberk JQJC, Amy GL, Van Dijk JC
(2012) Cation exchange during subsurface iron removal. Water
Res 46(2):307–315
WHO (2004) World Health Organization. Guidelines for Drinkingwater Quality 1:3
Yuce G (2007) Spatial distribution of groundwater pollution in the
Porsuk River Basin, Turkey. Environ Pollut 30:529–547
Yuce G, Pinarbasi A, Ozcelik S, Ugurluoglu D (2006) Soil and water
pollution derived from anthropogenic activities in the Porsuk
River Basin, Turkey. Environ Geol 49(3):359–375
123